Dockyai Chatgpt Sidebar Chrome Extension

What is Docky.AI?

Docky.AI is a browser AI sidebar operated by Thin Underline LLC; it was previously known as the Dockyai Chatgpt Sidebar Chrome Extension. It allows users to utilize various text and image models directly on the current webpage, thereby reducing the need to switch between the page and separate chat tools.

This product is not a single ChatGPT client; it also integrates Claude, Gemini, DeepSeek, Qwen, Grok, as well as various image models, and it distinguishes between usage levels based on basic and advanced credits.

Overview of main functions

Multi-model dialogue and group chats

Users can switch between models such as GPT, Claude, Gemini, and DeepSeek using the sidebar on the same side. The @ function allows users to pose questions to multiple models at once, so as to compare the structure, facts, and wording of their responses.

• Obtain different model-based explanation paths for the same research question.
• Have multiple models generate titles, outlines, or code solutions separately.
• Check whether the summary of a longer text omits any numbers, exceptions, and limitations.
• Choose a more suitable output for translation, rewriting, and creative tasks.
• In online search mode, issues that require up-to-date information are addressed.

Consistent answers from multiple models do not necessarily mean that the information is correct. In cases involving real-time events, professional advice, or important decisions, it is still necessary to consult the original sources and authoritative information.

Web reading and quick toolbar

Docky runs in the browser’s sidebar, and users can activate it on any webpage by pressing Cmd or Alt plus K. Once text on a page is selected, actions such as interpretation, translation, and continuation can be performed directly.

1. Install Docky from the browser extensions store, and check that the publisher is Thin Underline LLC.
2. Register or log in to your account, and select the free plan or a subscription as needed.
3. Open the webpage you want to work on, then use a shortcut key to bring up the sidebar.
4. Select a model and task, or click on text to activate the shortcut toolbar.
5. Check whether the output is faithful to the current page and add context.
6. After processing the sensitive pages, clear the chat and check the extension permissions.

The product can also display AI-generated responses alongside search results on Google, Bing, and other platforms. Search enhancements help reduce the number of steps required, but the generated summaries may combine the model’s existing knowledge with the content of the page.

PDF, Word, PPT, and image processing

File reading clearly supports PDF, Word, and PPT formats; it can be used to summarize content or to ask questions related to the files. The image function allows for the description of content, the identification of objects, and the extraction of text through OCR.

• First, confirm the number of pages, the language, and whether it is a scanned document.
• For long documents, first extract the table of contents and chapters, then ask questions section by section.
• The tables and figures must be checked item by item against the original text.
• The OCR results should be checked for spelling errors, line breaks, headers and footers, as well as handwritten content.
• Do not hand over documents containing passwords, identification documents, or customer confidential information to extensions directly.

The public page does not specify exact limits regarding file size, number of pages, OCR language, or retention period. The actual restrictions may vary depending on the account, model, and package used; it is necessary to follow the instructions provided within the product.

AI writing, responding, and translation

A writing assistant can generate articles, emails, social media posts, or replies based on tone, language, and length; it can also check for spelling and grammar errors. Once text is selected, it can be rewritten on the same page, without the need to copy it to another tab.

• Outline the draft using clear constraints regarding audience, purpose, tone, and length.
• It is required to retain facts, numbers, proper nouns, and the meaning of quotations.
• Replace sensitive personal information with placeholders before further processing.
• Check copyright, facts, and brand guidelines before releasing to the public.
• Manual review is carried out when translating important contracts or technical documents.

The content generated is at the user’s own discretion for use; the terms do not guarantee its accuracy. For marketing, academic, or commercial purposes, it is necessary to check for any repeated statements, factual errors, and unauthorized material.

Image generation and editing

Docky allows switching between various image models such as Flux, DALL-E, SDXL, and Ideogram; it can generate images based on prompts, dimensions, and style, or modify existing images. Different models may earn points at different levels.

The availability and names of the models may change depending on the supplier; therefore, the list of models provided on the pricing page should not be considered a permanent commitment. Before using them in commercial purposes, it is necessary to verify the copyright and terms of commercial use applicable to the selected models, packages, and the content that is generated.

Prices and packages

The free entry option indicates that no credit card is required, and the pricing page states that an annual payment can save 30% to 40%. The amount due for payment is not indicated on the public pages, and the names of the packages listed in the Chinese version may differ as well; therefore it is not possible to estimate the cost based on this information.

Mega’s “unlimited” plan still has limits regarding fair usage: once the monthly threshold for high-speed queries using advanced models is reached, the response speed and quality may decrease. Prices, taxes, currency types, available models, data quotas, and annual discount rates are specified on the account settlement page.

Refunds and package changes

The terms allow the service provider to adjust the pricing structure and billing methods; some free features may also be made payable, with notifications displayed on the page. The terms do not specify a clear deadline for refunds or rules regarding unused periods, so it is necessary to confirm the conditions for automatic renewal, cancellation, and refunds before subscribing.

Browser platforms and installation channels

The Chrome Store indicates that the extension supports 22 different interface languages, including Simplified Chinese, Traditional Chinese, English, Japanese, Korean, French, German, and others. The store mentions in-app purchases; free installation does not mean that all models and functions are available for free.

Expanded permissions and privacy

The Chrome Web Store states that these extensions manipulate website content, and they are related to functions such as page summarization, text highlighting, and search enhancements. Publishers claim that the data collected is not sold to third parties, nor is it used for purposes unrelated to the core functions of these extensions or for credit scoring.

The version date of the privacy policy is November 2023, and it specifies a wide range of data that is collected. This includes phone numbers, account details, device identifiers, network and location information, search terms, browsing history, service logs, content sent by users, as well as content generated by AI.

• The platform can examine user inputs, usage behaviors, and generated content to ensure compliance with regulatory requirements.
• The content may be submitted to a third party for review, and the results will be fed back to the platform.
• Anonymized information may be used for research, statistics, forecasting, product improvement, and training of model algorithms.
• The chat list provides a function to delete all chats; upon deletion, the relevant data will be removed or anonymized.
• Closing an account is irreversible; unless required by law, the account information will be deleted.
• Browser permissions, the clipboard, and page content can reveal sensitive information; therefore, the minimum necessary permissions should be used.

The information disclosed by a store differs from that provided in a comprehensive privacy policy, and the two cannot replace each other. Before using such systems, enterprises should verify the supplier of the modeling technology, the parties involved in data processing, any cross-border data transfers, the duration for which data is retained, as well as the procedures for excluding certain data from processing and for deleting it.

Terms, Copyrights, and Responsibility for Content

Users must ensure that the content they submit is covered by legitimate rights, and they are responsible for the consequences of using and sharing such content. The platform does not guarantee that the AI-generated output will be accurate, and it may review or remove any content that violates the rules, in accordance with the law.

• Content that is illegal, infringes on rights, involves fraud, harassment, or causes harm to others is prohibited.
• Reverse engineering, copying the system, altering its functions, or interfering with the service using unauthorized tools is prohibited.
• The intellectual property rights for the platform, interface, programs, and related content belong to the operator or its affiliated parties.
• Services may be upgraded, modified, interrupted, or discontinued; important data should be backed up separately.
• Upon termination of the service, the platform may permanently delete the data and is not obligated to restore it.

The terms do not clearly specify the uniform scope of authorization for using the generated images and text in commercial purposes. Commercial projects must also examine the licensing requirements related to Docky, the supplier of the selected models, the input materials, and the channels through which the content is published.

API, SDK, and open-source status

Suitable for users and typical use cases

• Study and understand users: Summarize, explain, and compare viewpoints in web pages and PDFs.
• Content creators: Use multiple models to generate outlines, rewrite text, and create illustrations.
• Office staff: Handling emails, Word documents, PPT presentations, as well as providing translation and grammar checks.
• Product and marketing staff: Improve search functions, organize competitor pages, and create communication materials.
• Power users of multiple models: Switch between models in a sidebar and compare answers in group chats.
• Users who frequently work across different web pages: use shortcut keys and word selection tools to reduce the need for copying and pasting.

Advantages and capabilities boundaries

Main advantages

• Aggregate text, files, pictures, and images in the browser sidebar.
• Multiple models are supported, and parallel comparison can be carried out via @ group chats.
• It can directly work on the current webpage, selected text, and search results.
• Supports OCR for PDF, Word, PPT, and images.
• Both Chrome and Edge have installation options available, and they support multiple interface languages.

Main limitations

• The model and prices change rapidly, and the public page does not display the current payment amount.
• The plan with unlimited credits may still experience speed reductions or a decline in quality when high usage is reached.
• Permissions for web page content can pose a risk of exposing sensitive data.
• The file size limit, retention period, and certain OCR boundaries are not clearly specified.
• The privacy policy is outdated and allows for content moderation as well as the use of anonymized data for training purposes.
• No public APIs, SDKs, native applications, or open-source code have been identified.

Frequently Asked Questions

Is Docky.AI free?

A free plan is available, offering 30 basic points per day; no credit card is required for registration. Advanced models and higher usage levels require a subscription.

Which browsers are supported?

Currently, Chrome and Microsoft Edge are confirmed to work. It is necessary to test on your own to determine whether other Chromium browsers can be installed and updated stably.

Can multiple models be queried at the same time?

Yes. By using @ group chats, you can send the same question to multiple models and compare their answers in the sidebar.

Can you summarize the current webpage and PDF?

Yes, it also supports Word, PPT, as well as image recognition and OCR. Long documents and complex tables still need to be checked against the original text.

Can all models be used indefinitely after free installation?

No. The free version uses basic points on a daily basis, and the paid versions also have separate basic and advanced point systems.

Are Mega’s advanced models truly unlimited?

The page states that there is no limit on the number of queries, but it guarantees at least 600 high-speed queries to the advanced models per month. Once these limits are exceeded, the speed and quality of the queries may decline.

Does Docky read web page content?

The website content is processed to enable summarization, text highlighting, and improved search functionality. When internal systems, customer data, or personal information are involved, the extent of access should be restricted or use should be suspended.

Can I delete chat data?

The privacy policy states that the chat list offers a function to delete all chats; the relevant data is deleted or anonymized, unless legal requirements dictate that it be retained.

Is Docky.AI open source?

It is not an open-source product that can be verified. The presence of a browser extension, a list of models, or a listing in a store does not mean that the source code of the product is available publicly.

Are there APIs or SDKs?

Currently, no official APIs or SDKs available for ordinary developers exist. When enterprise integration is required, it is necessary to consult the operators directly.

Can the generated content be used for commercial purposes?

The public terms do not provide a unified commercial license that covers all models and materials. Before publishing, it is necessary to check the rules regarding the selected models, input materials, account packages, and the region of use.

Summary

Docky.AI is suitable for users who wish to carry out multi-model Q&A, read web pages and files, perform writing and translation tasks, as well as generate images within Chrome or Edge. Before choosing a plan, it is important to verify the actual pricing, the amount of model usage allowed, the permissions granted by the browser, and the scope of content processing.